Journal of Fish Biology (2009) 75, 2326–2343doi:10.1111/j.1095-8649.2009.02489.x, available online at www.interscience.wil...
PHYLOGEOGRAPHY OF HOPLIAS MALABARICUS                                                         2327information relevant for...
2328                                             U. SANTOS ET AL.                                   MATERIALS AND METHODSS...
PHYLOGEOGRAPHY OF HOPLIAS MALABARICUS                                                          2329 Equator               ...
2330                                             U. SANTOS ET AL.cycle sequencing kit (PE Applied Biosystems; www.appliedb...
PHYLOGEOGRAPHY OF HOPLIAS MALABARICUS                                                       2331(a)                       ...
2332                                              U. SANTOS ET AL.          (a)                               (b)         ...
PHYLOGEOGRAPHY OF HOPLIAS MALABARICUS                                                         2333           (a)          ...
2334                                             U. SANTOS ET AL.                   (a)                           m       ...
PHYLOGEOGRAPHY OF HOPLIAS MALABARICUS                                                       2335                          ...
2336                                             U. SANTOS ET AL.(Born & Bertollo, 2000). The 2n = 40/41G karyomorph also ...
PHYLOGEOGRAPHY OF HOPLIAS MALABARICUS                                                         2337the 2n = 42B karyomorph ...
2338                                             U. SANTOS ET AL.and Grande Rivers. Some 2n = 42 trahiras from the Tombado...
PHYLOGEOGRAPHY OF HOPLIAS MALABARICUS                                                         2339involved few alterations...
2340                                             U. SANTOS ET AL.its associated haplotypes were widely distributed through...
PHYLOGEOGRAPHY OF HOPLIAS MALABARICUS                                                         2341and Itapicuru River, and...
2342                                             U. SANTOS ET AL.       with emphasis on the Hoplias malabaricus ‘species ...
PHYLOGEOGRAPHY OF HOPLIAS MALABARICUS                                                         2343Quenouille, B., Bermingh...
Upcoming SlideShare
Loading in …5

Molecular and cytogenetic phylogeography of h. malabaricus


Published on

Claudio Michael Völcker
Jorge A. Dergam

Molecular and karyotypic phylogeography in the Neotropical Hoplias malabaricus (Erythrinidae) fish in eastern Brazil

Published in: Technology
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Molecular and cytogenetic phylogeography of h. malabaricus

  1. 1. Journal of Fish Biology (2009) 75, 2326–2343doi:10.1111/j.1095-8649.2009.02489.x, available online at Molecular and karyotypic phylogeography in the Neotropical Hoplias malabaricus (Erythrinidae) fish in eastern Brazil ¨ U. Santos*, C. M. Volcker†, F. A. Belei*, M. B. Cioffi‡, L. A. C. Bertollo‡, S. R. Paiva§ and J. A. Dergam* *Laborat´ rio de Sistem´ tica Molecular Beagle, Departamento de Biologia Animal, o a Universidade Federal de Vi¸ osa, 36570-000 Vi¸ osa, MG, Brazil, †Rua Edgard Coutens de c c Menezes, 47. 28970-000 Pontinha, Araruama, RJ, Brazil, ‡Centro de Ciˆ ncias Biol´ gicas e o e da Sa´ de, Departamento de Gen´ tica e Evolu¸ ao, Universidade Federal de S˜ o Carlos, u e c˜ a S˜ o Carlos, SP, Brazil and §Embrapa Recursos Gen´ ticos e Biotecnologia, Laborat´ rio a e oGen´ tica Animal, PCG, Parque Esta¸ ao Biol´ gica, Final Av. W/5 Norte, Bras´lia, DF, Brazil e c˜ o ı (Received 9 June 2009, Accepted 13 October 2009) The sedentary, predatory characin Hoplias malabaricus has one of the widest distributions of fresh- water fishes in South America and is characterized by seven karyomorphs (A–G) that occur in sympatric and allopatric populations. Karyotypical patterns of variation in wild populations have been interpreted as evidence of multiple lineages within this nominal species, a possibility that may limit the validity of experimental data for particular karyomorphs. This study used the phylogeo- graphic and genealogical concordance between cytogenetic (N = 49) and molecular (mitochondrial DNA) (N = 73) data on 17 samples, collected in 12 basins from south-eastern and north-eastern Brazil, to assess the systematic value of cytogenetic data. Cytogenetic patterns show a sex chro- mosome system in the 2n = 40F karyomorph. Molecular and cytogenetic data indicate a long, independent evolutionary history of karyomorphs and a coastal origin of continental populations in south-eastern Brazil. The lack of fit with molecular clock expectations of divergence between groups is likely to be due to strong demographic fluctuations during the evolution of this species complex. The results indicate that karyotypical identification provides a reliable baseline for placing experimental studies on Hoplias spp. in a phylogenetic context. © 2009 The Authors Journal compilation © 2009 The Fisheries Society of the British Isles Key words: freshwater fish; in situ hybridization; molecular clock; mtDNA; repetitive DNAs; South America. INTRODUCTIONThe Neotropical region harbours at least 6025 species of freshwater fishes (Reiset al., 2003), but perhaps as many as 8000 (Schaefer, 1998), the result of a long evo-lutionary history of isolation and specialization, involving mainly otophysan fishes.Freshwater fishes are highly suitable for the recovery of past biogeographic processes,due to their obligatory relationship with water (Myers, 1938). The biogeographical Author to whom correspondence should be addressed. Tel.: +55 31 3899 2555; fax: +55 31 3899 2578;email: 2326 © 2009 The Authors Journal compilation © 2009 The Fisheries Society of the British Isles
  2. 2. PHYLOGEOGRAPHY OF HOPLIAS MALABARICUS 2327information relevant for fish diversification has been reviewed by Lundberg et al.(1998) and Ribeiro (2006). The former revision focused on the evolution of Andeangeomorphology and the effects of this process on the South American continent,whereas the latter work dealt with the evolution of particular coastal drainages,their geomorphological interactions with continental basins and the expected lev-els of phylogenetic divergence among fishes derived during these complex temporalevents. Among Neotropical freshwater fishes, the trahira, Hoplias malabaricus (Bloch),has one of the largest distributional ranges, occurring from Panam´ to the Buenos aAires Province in Argentina (Berra, 2007). H. malabaricus is well adapted to lifein small, isolated populations, in conditions that may facilitate the stochastic fixa-tion of chromosomal rearrangements (Sites & Moritz, 1987). This species is one ofthe cytogenetically most studied taxa and shows a conspicuous karyotypic diversi-fication, with up to seven karyomorphs with diploid numbers ranging from 39 to42. Three groups of karyomorphs are readily evident, based on distribution ranges.Two karyomorphs, 2n = 42A and 2n = 40C, are the most widespread and occur insympatry in the lower Paran´ –Paraguay and Amazon River basins. A second, less awidely spread group is represented by three karyomorphs: (1) 2n = 40F occurs inthe S˜ o Francisco River basin, in the lower portions of the Tocantins River, in coastal adrainages in Surinam and in minor basins in north-eastern Brazil; (2) 2n = 39/40Doccurs in the upper Paran´ River (in sympatry with 2n = 42A) and (3) 2n = 40/41G aoccurs exclusively in the Amazonian Basin in the Aripuan˜ , Madeira and Trombetas aRivers. The third group consists of karyomorphs with much restricted distributions. Forexample, 2n = 42B occurs in the Doce River basin (also in sympatry with2n = 42A), and 2n = 42E occurs only in the Trombetas River (Bertollo et al., 2000). The lack of hybrids between sympatric karyomorphs has been interpreted as evi-dence for the existence of several distinct species within this nominal taxon (Bertolloet al., 1986, 2000). Therefore, uniparental molecular markers can be a particularlyuseful complement to assess gene flow in wild populations. To date, molecular stud-ies on H. malabaricus have been restricted to populations in the Doce River basin(Dergam et al., 2002) and the Iguacu River basin (Dergam et al., 1998). Both studies ¸indicated phylogenetically related populations in the coastal and continental basins,but failed to reveal the direction of dispersal events. Karyotypic and molecular patterns of variation may reveal vicariances and dis-persals that provide insights into evolutionary processes that may involve othercomponents of the aquatic fauna. The congruence of independent molecular andcytogenetic characters may give support to the hypothesis that at least some kary-omorphs behave as valid species within H. malabaricus. A better understanding ofH. malabaricus species-level systematics is needed to interpret correctly the compar-ative physiological, behavioural and anatomical studies in an explicit phylogeneticframework (Harvey & Pagel, 1991). Biodiversity patterns in this species complexalso bear potential information to understand the palaeohydrology of the continent. Inthe present study, patterns of cytogenetic and molecular variation of H. malabaricuspopulations in the S˜ o Francisco, Grande and in several coastal basins along eastern aBrazil were analysed using molecular and cytogenetic techniques.© 2009 The AuthorsJournal compilation © 2009 The Fisheries Society of the British Isles, Journal of Fish Biology 2009, 75, 2326–2343
  3. 3. 2328 U. SANTOS ET AL. MATERIALS AND METHODSS P E C I M E N S A N D C H R O M O S O M E P R E PA R AT I O N Cytogenetic analyses were carried out on 49 specimens collected in the Pandeiros River,Par´ River and its headwaters in the Tombadouro Creek and in the Jacar´ River (Table I a eand Fig. 1). Cell division was stimulated in vivo with two daily applications of Munolan, acommercially available antigen lysate, following Molina (2001). Mitotic chromosomes wereobtained from cell suspensions of the anterior kidney, using the conventional air-dryingmethod (Bertollo et al., 1978). Fish were previously anaesthetized with clove oil (Henyeyet al., 2002). Secondary data from 10 populations with known chromosome numbers werealso included in the analysis.C H R O M O S O M E S TA I N I N G In addition to the standard Giemsa method, chromosomes were analysed using silver nitratestaining (Howell & Black, 1980) to visualize the nucleolar organizing regions (Ag-NOR).C-banding was also employed to detect C-positive heterochromatin (Sumner, 1972). (4 ,6diamidino-2-phenylindole dihydrochloride (DAPI) and chromomycin A3 (CMA3 ) fluores-cence staining were used to identify the chromosome adenine–thymine (AT) and guanine-cytosine (GC) rich regions, respectively (Sola et al., 1992). Some populations with unknownkaryotypes (or whose karyotypes are being studied by J.A.D.) were also included in this study(see below).C H R O M O S O M E H Y B R I D I Z AT I O N A N D K A RY O T Y P I CA N A LY S I S Fluorescent in situ hybridization (FISH) was performed according to Pinkel et al. (1986)using three repetitive DNA sequences probes isolated from the genome of H. malabaricus andan 18S probe. The first probe contained a 5S rDNA repeat copy and included 120 base pairsTable I. Hoplias malabaricus collecting localities, sample sizes, geographic coordinates and karyomorph nomenclature Cytogenetic MolecularLocality samples samples G.P.S. KaryomorphPar´ River a 8♀ –7♂ 14 20◦ 08 21 S–44◦ 53 17 W 40FPandeiros River 2♀ –3♂ 9 15◦ 40 18 S–44◦ 37 43 W 40FTombadouro Creek 4♀ –3♂ 5 20◦ 41 56 S–44◦ 34 46 W 42AIbimirim — 6 8◦ 30 31 S–37◦ 42 21 W UnknownCurvelo 12 8 18◦ 42 57 S–44◦ 25 56 W 40FItapicuru River 4 1 13◦ 14 40 S–41◦ 23 34 W 40FDom Helv´ cio Lake e 10 3 19◦ 46 34 S–42◦ 35 19 W 42BS˜ o Mateus River a 2 6 18◦ 44 09 S–48◦ 29 47 W 42APara´ba do Sul River ı 15 2 21◦ 46 26 S–41◦ 30 01 W 42AItabapoana River 8 2 21◦ 08 20 S–41◦ 39 35 W 42AMaca´ River e 6 1 22◦ 17 43 S–41◦ 52 48 W 42AS˜ o Jo˜ o River a a 8 2 22◦ 33 29 S–42◦ 07 21 W 42ARibeira River 2 1 24◦ 29 23 S–47◦ 50 10 W 42ACarioca Lake 15 1 19◦ 45 32 S–42◦ 37 15 W 42BParanagu´ Bay a — 1 25◦ 32 39 S–48◦ 29 47 W UnknownMacacu Waterfalls — 1 22◦ 27 37 S–42◦ 39 18 W UnknownJacar´ River e 15♀ –7♂ 10 20◦ 48 29 S–44◦ 33 58 W 42A © 2009 The Authors Journal compilation © 2009 The Fisheries Society of the British Isles, Journal of Fish Biology 2009, 75, 2326–2343
  4. 4. PHYLOGEOGRAPHY OF HOPLIAS MALABARICUS 2329 Equator 35° 44′ 58′′ W 8° 08′ 01′′ S 1 500 km 1 - Ibimirim 2 2 - Itapicuru River 3 - Pandeiros River 3 4 - Buranhém River 5 - Curvelo 6 - Três Marias 7 - São Mateus River 4 8 - Dom Helvécio and Carioca lakes 5 9 - Pará River 6 7 10 - Tombadouro Creek 8 9 11 - Itabapoana 10 12 11 12 - Jacaré River 13 13 - Paraíba do Sul River 14 14 - Macacu 16 15 15 - Macaé 17 18 16 - São João River 26° 19′ 17′′ S 17 - Ribeira River 54° 35′ 13′′ W 18 - ParanaguáFig. 1. Hoplias malabaricus collection localities. The dotted line indicates the upper Grande crustal disconti- nuity. The locality of Trˆ s Marias is also depicted. e(bp) of the 5S rRNA encoding gene and 200 bp of the non-transcribed spacer (NTS) (Martinset al., 2006). The second probe is specific to H. malabaricus (Ferreira et al., 2007) andcontained a copy of the repetitive satellite 5SHind III-DNA sequence with 360 bp composedof a 95 bp segment similar to the 5S rRNA gene of the first probe and a 265 bp segment similarto the NTS of the first probe (Martins et al., 2006). The third probe corresponded to a 1400 bpsegment of the 18S rRNA gene obtained by the polymerase chain reaction (PCR) fromnuclear DNA (Cioffi et al., 2009). The probes were labelled by nick translation with biotin-14-dATP (Bionick Labeling System, Invitrogen; Signal detection and itsamplification were performed using conjugated avidin–fluorescein isothiocyanate (FITC) andanti-avidin–biotin (Sigma; The chromosomes were counterstainedwith propidium iodide (50 μg ml−1 ) and analysed with an Olympus BX50 epifluorescencemicroscope. The chromosomal images were captured using CoolSNAP-Pro software (MediaCybernetics; About 30 metaphase spreads were analysed per specimento determine the diploid chromosome number and karyotypic structure. Chromosomes wereclassified as metacentrics (m) or submetacentrics (sm), according to centromeric index valuesproposed by Levan et al. (1964).D N A E X T R A C T I O N A N D M O L E C U L A R A N A LY S I S DNA extraction followed Boyce et al. (1989). Fragments of ATPase 6 were amplifiedusing primers L8524 (5 -AAY CCT GAR ACT GAC CAT G-3 ) and H9236 (5 -GTT AGTGGT CAK GGG CTT GGR TC-3 ) (Quenouille et al., 2004). Double-stranded DNA was syn-thesized in 50 μl reactions containing 10 μl dNTPs (1 mM), 5 μl reaction buffer (200 mMTris–HCl pH 8·4, 500 mM KCl), 2 μl MgCl2 (50 mM), 2 μl of each primer (10 mM), 0·5 μl(2·5 U) Taq DNA polymerase (Phoneutria), 2 μl template DNA (100 ng/μl) and 26·5 μl H2 O.PCR conditions were as follows: 94◦ C (2 min), five cycles of 94◦ C (45 s), 54◦ C (45 s) and72◦ C (1·5 min) and 29 cycles of 94◦ C (45 s), 58◦ C (45 s) and 72◦ C (1·5 min). PCRproducts were purified using Qiaquick (Qiagen; 5 μl of the purifiedPCR product were used in a 10 μl cycle sequencing reaction using a drhodamine terminator© 2009 The AuthorsJournal compilation © 2009 The Fisheries Society of the British Isles, Journal of Fish Biology 2009, 75, 2326–2343
  5. 5. 2330 U. SANTOS ET AL.cycle sequencing kit (PE Applied Biosystems; Sequences werealigned using CLUSTAL X 1.83 in MEGA 4.0 (Tamura et al., 2007). Phylogenetic trees wereconstructed with neighbour joining (NJ) (Saitou & Nei, 1987), maximum parsimony (MP),maximum likelihood (ML) (Felsenstein, 1981) and Bayesian inference (MB) (Huelsenbeck &Ronquist, 2001). The model of molecular evolution that best fitted the data was chosen usingModeltest 3.7 (Posada & Crandall, 1998), and haplotype divergence was estimated withinand between haplogroups using this selected model. Phylogenetic signal in NJ, MP and MLtrees was assessed using bootstraps with 1000 repetitions. Bayesian inference was performedwith five million Markov-Chain Monte-Carlo (MCMC) steps to produce posterior probabil-ities of nodes in the tree with MrModeltest 2 (Nylander, 2004). Hoplias lacerdae MirandaRibeiro was used as an outgroup. Sequences were deposited in GenBank (accession numbersGQ848606–GQ848642). RESULTSK A RY O T Y P I C D ATAKaryomorph 2n = 40F All specimens from the Pandeiros and Par´ rivers (except those from Par´ ’s a aheadwaters, in Tombadouro Creek) had 2n = 40 chromosomes for both sexes. Thekaryotypes were composed of 10 m + 10 sm pairs, without morphologically differ-entiated sex chromosomes using Giemsa conventional staining [Fig. 2(a)], and weretypical 2n = 40F karyomorphs. C-banding showed heterochromatic blocks in cen-tromeric regions, except for chromosome pairs 9 and 10, which had less conspicuousblocks, in addition to faint telomeric marks in some chromosome pairs. Males alwaysshowed an interstitial heterochromatic block in the short arms of one homologue ofthe first chromosome pair [Fig. 2(c)], whereas females lacked this block [Fig. 2(e)].This heterochromatic region was negative for DAPI staining but failed to showfluorescence with CMA3 . A proximal heterochromatic block located on the shortarms of a small submetacentric pair was the only observed GC-rich segment in bothpopulations [Fig. 3(f), (i)]. Ag-NORs were telomeric and their numbers were eitherfixed (four in the Par´ River) or varied from 4 to 5 (Pandeiros River) [Fig. 4(a), (b)]. aFISH analyses with the repetitive sequences showed no differences between these twopopulations. The 18S rDNA probe hybridized on the pericentromeric region of onemetacentric pair and on the telomeric regions of two submetacentric pairs [Fig. 3(a)].The 5S rDNA probe hybridized on the pericentromeric region of a medium-sizedmetacentric pair [Fig. 3(c)]. The repetitive 5SHind III-DNA was mapped in the cen-tromeric region of 10 chromosome pairs [Fig. 5(a)]. Although Ibimirim samples werenot karyotyped, they were collected in a region with populations characterized by2n = 40F karyomorph and were assumed to share this karyomorph.Karyomorph 2n = 42A In the headwaters of Tombadouro Creek, all specimens had 2n = 42 chromosomes,with a karyotype consisting of 11 m + 10 sm pairs and without morphologically dif-ferentiated sex chromosomes [Fig. 2(b)], and were considered to be typical 2n = 42Akaryomorphs. Besides some telomeric marks, conspicuous heterochromatic bandswere found in the centromeric region of all chromosomes, except for chromosomepairs 6, 19 and 20, which showed faint blocks [Fig. 2(d)]. Ag-NORs varied amongand within individuals (one to two pairs), and bitelomeric NORs were evident ineither one or two chromosomes [Fig. 4(c)]. In this population, the 18S rDNA probe © 2009 The Authors Journal compilation © 2009 The Fisheries Society of the British Isles, Journal of Fish Biology 2009, 75, 2326–2343
  6. 6. PHYLOGEOGRAPHY OF HOPLIAS MALABARICUS 2331(a) (b) m m 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9 10 10 11 sm sm 11 12 13 14 15 16 17 18 19 12 13 14 15 16 17 18 19 20 20 21(c) (d) m 1 2 3 4 5 6 7 8 9 m 1 2 3 4 5 6 7 8 9 10 10 11 sm sm 11 12 13 14 15 16 17 18 19 12 13 14 15 16 17 18 19 20 20 21(e) (f) m m 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9 10 10 11 sm sm 11 12 13 14 15 16 17 18 19 12 13 14 15 16 17 18 19 20 20 21 5 μmFig. 2. Giemsa-stained and C-banded karyotypes of Hoplias malabaricus. (a) conventional Giemsa-stained karyotype of populations from Pandeiros and Par´ rivers, (b) samples from Jacar´ River and Tombadouro a e Creek, (c) C-banded karyotypes of males from Par´ and Pandeiros Rivers, (e) C-banded karyotypes of a females from Pandeiros and Par´ Rivers, (d) C-banded karyotype of specimens from Tombadouro Creek a and (f) C-banded karyotype from Grande River. Arrows indicate an interstitial heterochromatic block in the short arm of a homologue of the first chromosome pair. This block appeared in males of Pandeiros and Par´ River populations, but not in females. ahybridized on three chromosome pairs, with interstitial, telomeric and bitelomericsites [Fig. 3(b)]. The 5S rDNA probe hybridized on the interstitial region of a largesubmetacentric pair [Fig. 3(d)]. The repetitive 5SHind III-DNA probe bound to thecentromeric region of nine chromosome pairs [Fig. 5(b)]. Specimens from the Jacar´ River in the Grande drainage had 2n = 42 kary- eomorphs composed of 11 m + 10 sm pairs, without morphologically differentiatedsex chromosomes [Fig. 2(b)], and represented a typical 2n = 42A karyomorph. Het-erochromatic blocks were centromeric in most chromosomes, except for pairs 6,8, 11, 19 and 20, which showed faint bands, in addition to telomeric marks insome chromosome pairs [Fig. 2(f)]. Ag-NORs were restricted to telomeric regionsof two chromosomes [Fig. 4(d)]. Fluorescent regions hybridized with 18S rDNAand 5SHind III-DNA and showed a pattern similar to the Tombadouro Creek sample.© 2009 The AuthorsJournal compilation © 2009 The Fisheries Society of the British Isles, Journal of Fish Biology 2009, 75, 2326–2343
  7. 7. 2332 U. SANTOS ET AL. (a) (b) (c) (d) (e) (f) (g) (h) (i)Fig. 3. Metaphase chromosome spreads of Hoplias malabaricus after FISH with 18S and 5S rDNA probes and DAPI–Chromomycin A3 staining. Numerical and positional variation of 18S rDNA sites in populations from (a) Pandeiros and Par´ Rivers, (b) Tombadouro Creek and Jacar´ River. Mapping of 5S rDNA sites a e for populations of (c) Pandeiros and Par´ Rivers, (d) Tombadouro Creek and (e) Jacar´ River. DAPI a e staining demonstrating absence of signals in males (f) and females (h) from Par´ River. Arrow indicates a the interstitial heterochromatic block negative for DAPI in one homologue of the first chromosome pair in males. Sequential CMA3 staining showing GC-rich DNA segments located on a submetacentric chromosome pair on (g) males and (i) females. Bar = 5 μm.Additionally, the 5S rDNA probe hybridized on the interstitial region of a smallmetacentric pair [Fig. 3(e)]. As was the case with the Ibimirim samples, Paranagu´ aand Macacu samples were collected within the range of 2n = 42A populations(J. A. Dergam, unpubl. data) and were therefore assumed to be derived from thesepopulations.M O L E C U L A R D ATA Multiple sequences 511 bp in length yielded 86 phylogenetically informative sites.Translation of these sequences into the corresponding amino acid sequences resultedin eight phylogenetically informative sites. The transitions:transversions ratio was6·4, indicating that substitution rates were not saturated. The best model fit esti-mated with Modeltest and MrModeltest was HKY+G, which was used for NJ.In the tree, haplotypes were separated into two large clades with variable boot-strap support (Fig. 6). The largest bootstrap values were obtained for haplogroup I,which was composed of the 2n = 42A karyomorph from Tombadouro Creek, Ribeira,Paranagu´ and Macacu rivers (Fig. 1). Haplogroup II was less supported and included aall remaining H. malabaricus (N = 67), regardless of diploid number. Within this © 2009 The Authors Journal compilation © 2009 The Fisheries Society of the British Isles, Journal of Fish Biology 2009, 75, 2326–2343
  8. 8. PHYLOGEOGRAPHY OF HOPLIAS MALABARICUS 2333 (a) (b) 5 μm 5 μm (c) (d) 5 μm 5 μmFig. 4. Metaphase chromosome spreads of Hoplias malabaricus after silver staining, showing Ag-NORs in populations from (a) Par´ River, (b) Pandeiros River, (c) Tombadouro Creek and (d) Jacar´ River. a e Arrows and arrowhead indicate telomeric and bitelomeric Ag-NORs, respectively. Table II. Within- and between-haplogroup molecular distances in Hoplias malabaricus Haplogroup I, Haplogroup IIA, Haplogroup IIB, 2n = 42 2n = 40 2n = 42Haplogroup I, 2n = 42 0·032Haplogroup II A, 2n = 40 6·82 0·0053Haplogroup II B, 2n = 42 9·20 9·50 2·65haplogroup, two subclades were highly supported: haplotypes derived from popula-tions with the 2n = 40F karyomorph (haplogroup IIA) and from populations with2n = 42A and 2n = 42B karyomorphs (haplogroup IIB). Haplogroup IIA includedall 2n = 40 S˜ o Francisco River H. malabaricus, plus the Itapicuru River specimen. aHaplogroup IIB included two well-defined lineages: one included haplotypes fromthe Doce and S˜ o Mateus Rivers with no representatives in the Grande River, and aa second lineage indicated a close phylogenetic relatedness between H. malabaricusfrom the Jacar´ (Grande River), the Tombadouro (S˜ o Francisco River) and the e aMaca´ , S˜ o Jo˜ o, Itabapoana and Para´ba do Sul coastal basins. Within-haplogroup e a a ınucleotide molecular distances showed high levels of variation, with the smallest val-ues in the S˜ o Francisco and Itapicuru samples and the largest in the IIB haplogroup. aThe range of between-haplogroup variation was much smaller (Table II).© 2009 The AuthorsJournal compilation © 2009 The Fisheries Society of the British Isles, Journal of Fish Biology 2009, 75, 2326–2343
  9. 9. 2334 U. SANTOS ET AL. (a) m 1 2 3 4 5 6 7 8 9 10 sm 11 12 13 14 15 16 17 18 19 20 (b) m 1 2 3 4 5 6 7 8 9 10 11 sm 12 13 14 15 16 17 18 19 20 21Fig. 5. Karyotypes of Hoplias malabaricus samples from (a) Par´ and Pandeiros Rivers and (b) Tombadouro a Creek and Jacar´ River arranged from chromosomes probed with 5SHind III-DNA counterstained with e propidium iodide. Bar = 5 μm. DISCUSSIONK A RY O T Y P I C D ATAKaryomorph 2n = 40F In addition to specimens from Tombadouro Creek, all samples from the S˜ o Fran- acisco drainage shared the same karyomorphic formula, which is considered to becharacteristic for S˜ o Francisco basin populations (Bertollo et al., 2000). The dis- atributional pattern of heterochromatin was similar to that described by Dergam &Bertollo (1990) for the Trˆ s Marias population in this basin (Fig. 1), except for the epresence of a heterochromatic block that was always restricted to one homologueof the first chromosome pair in males. This pattern suggested a probable XX/XYsex chromosome system in these populations, because this variant occurred onlyin males and was absent in females. This sex chromosome differentiation involv-ing only heterochromatinization is a novelty within H. malabaricus, because allother sex chromosome systems described for this species complex involve transloca-tions between chromosome pairs (e.g. karyomorphs 2n = 39/40D and 2n = 40/41G)(Bertollo et al., 2000), or translocation and heterochromatinization, as proposed forthe evolution of the 2n = 42B karyomorph from a 2n = 42A karyomorph ancestor © 2009 The Authors Journal compilation © 2009 The Fisheries Society of the British Isles, Journal of Fish Biology 2009, 75, 2326–2343
  10. 10. PHYLOGEOGRAPHY OF HOPLIAS MALABARICUS 2335 40F Pará River 40F Pandeiros River Ibimirim Haplogroup II A 40F Pandeiros River 40F Itapicuru River 40F Curvelo River 100, 1·00, 100, 100 Ibimirim 40F Pandeiros River 40F Pará River 40F Curvelo 40F Pandeiros River 66, 1·00, 83, 71 42B Carioca Lake 42A São Mateys River 42B Dom Helvécio Lake 42 A São Mateus River 42A São Mateus River 100, 1·00, 98, 98 42B Dom Helvécio Lake 42A São Mateus River 42A Macaé River 63, 1·00, 86, 64 42A São João River 62, 0·97, *, 61 Haplogroup II B 42 A Paraíba do Sul River 42A Itabapoana River 99, 1·00, 90, 100 42A Paraiba do Sul River 42A Tombadouro Creek 42A Jacaré River Macacu 100, 1·00, 83, 100 Paranaguá 42A Ribeira River Haplogroup I 42A Tombadouro Creek Hoplias lacerdae 0·02Fig. 6. Phylogenetic relationships of ATPase 6 haplotypes derived from Hoplias malabaricus karyomorphs. The topology was obtained with neighbour joining analysis. Bootstrap values are expressed as neighbour joining–Bayesian–maximum likelihood–maximum parsimony. The asterisk indicates a polytomy with maximum likelihood analysis. Bar = molecular distance.© 2009 The AuthorsJournal compilation © 2009 The Fisheries Society of the British Isles, Journal of Fish Biology 2009, 75, 2326–2343
  11. 11. 2336 U. SANTOS ET AL.(Born & Bertollo, 2000). The 2n = 40/41G karyomorph also has a large first pair ofmetacentrics and is therefore morphologically closest to the 2n = 40F karyomorph.Females with both karyomorphs have identical karyotypes and may be sister kary-omorphs or species. Ongoing studies including these karyomorphs, using additionalchromosomal markers will allow for a more thorough hypothesis of the evolutionaryorigins of this sex chromosome system. C-positive heterochromatic bands were always located in the centromeric/pericentromeric region of all chromosomes and in the telomeric region of somepairs, in addition to multiple Ag-NORs sites. These heterochromatic and NOR char-acteristics were similar to those reported for other populations or karyomorphs ofH. malabaricus (Dergam & Bertollo, 1990; Haaf et al., 1993; Bertollo, 1996, 1997;Born & Bertollo, 2000; Vicari et al., 2003, 2005). In the Pandeiros River sample,Ag-NORs were restricted to telomeres of some chromosomes and appeared in largernumbers than 18S rDNA sites, but the latter also hybridized to the pericentromericregion of one chromosome pair [Fig. 3(a)]. There is currently no explanation for theapparently non-specific nature of some telomeric Ag-NORs in the Pandeiros sample,which is as large as seven in the Trˆ s Marias region (Dergam & Bertollo, 1990). On ethe other hand, pericentromeric 18S rDNA sites were not visualized by the silvernitrate method (these data and Dergam & Bertollo, 1990), which may be due topreferential telomeric NOR activation, as already pointed out for other Hoplias spp.karyomorphs (Vicari et al., 2005). Hybridization sites of the 5S rDNA probe wereinterstitially located in a small metacentric pair in both populations, a pattern alsoreported for H. malabaricus samples from Trˆ s Marias (Ferreira et al., 2007). The epatterns obtained with repetitive DNA class 5SHind III-DNA were similar to thoseobtained for the Trˆ s Marias population (Ferreira et al., 2007). eKaryomorph 2n = 42A The karyotypes of specimens from Tombadouro Creek and Jacar´ River were char- eacterized as karyomorph 2n = 42A, a widespread karyotypic form in the Neotropics(Bertollo et al., 2000). C-banding slowed conspicuous centromeric blocks in almostall chromosomes, an overall pattern also reported by Born & Bertollo (2001) for otherGrande River 2n = 42A populations. High levels of Ag-NOR variation observed inthe Tombadouro Creek population have also been reported elsewhere in the GrandeRiver basin (Born & Bertollo, 2001). Jacar´ River H. malabaricus showed Ag-NORs erestricted to one chromosome pair, characterizing the lowest number of activatedNORs reported so far as a fixed character state for any H. malabaricus population.FISH with an 18S rDNA probe, however, showed a larger number of NORs cistronsin both 2n = 42A karyomorph populations. On the other hand, FISH mapping with a 5S rDNA probe revealed differencesbetween the Tombadouro Creek and the Jacar´ River populations, where they ehybridized on a large submetacentric and a small metacentric pair, respectively.This difference contrasted with previous reports of other H. malabaricus popula-tions from the upper Paran´ Basin, the Aragu´ River, where both pairs showed a afluorescence with this probe (Martins et al., 2006). Considering that at least somespecimens from both basins shared identical haplotypes (see below), these populationdifferences highlighted the great potential of this probe for population studies. Patterns of variation of 18S rDNA were substantially different from the onesreported by Cioffi et al. (2009). These authors indicated the presence of five sites in © 2009 The Authors Journal compilation © 2009 The Fisheries Society of the British Isles, Journal of Fish Biology 2009, 75, 2326–2343
  12. 12. PHYLOGEOGRAPHY OF HOPLIAS MALABARICUS 2337the 2n = 42B karyomorph and four fluorescent sites in a 2n = 42A karyomorph fromthe upper Paran´ River basin, whereas the results of this study showed that the same akaryomorph in the Jacar´ and Tombadouro populations had three fluorescent sites. eThis reduction in numbers was also evident in the 2n = 40F karyomorph, whencompared with the 2n = 40C and 2n = 39/40D karyomorphs, which showed fivefluorescent sites. Thus, this probe appears to be especially informative for studyingamong-karyomorph variation. According to the haplotype-based trees, 5SHind III-DNA patterns reflected a smallincrease in site numbers (from 18 to 20); Cioffi et al. (2009) also found variationin site numbers for other 2n = 42 and 2n = 40 populations. Also, the fluorescentcentromeric region in the first or second chromosome pairs appears to be particularlyinformative. The presence of the fluorescent region in the first chromosome pair wasobserved in the 2n = 42A karyomorph, as is the case for 2n = 42B, 2n = 40C and2n = 39/40D karyomorphs (Cioffi et al., 2009), representing a plesiomorphic state.On the other hand, the 2n = 40F karyomorph showed an apparently apomorphiccondition, where the fluorescent region occurred in the second chromosome pair.Absence of the fluorescent site in the first chromosome pair might be attributedto a more recent origin of the first chromosome pair and its possible origin fromchromosomes lacking this site. This hypothesis, however, involves a particularlycomplex evolutionary process, because the persistence of diploid number would alsoinvolve alterations of other chromosome pairs. Alternatively, the 2n = 40F kary-omorph might have evolved directly from a 2n = 42A ancestor, causing a reductionin diploid number and producing an independent 2n = 40 lineage. This hypothesis isconsistent with the hypothesis of Bertollo et al. (2000), based on gross chromosomemorphology, that the 2n = 40F, 2n = 40/41G and 2n = 42E karyomorphs representa monophyletic group within H. malabaricus. In the submetacentric chromosome group, the pattern of fluorescent centromericsites in three major chromosome pairs also sets apart the 2n = 40F karyomorph fromother karyomorphs. The 2n = 42A karyomorph and the three karyomorphs reportedby Cioffi et al. (2009) have conspicuous fluorescent centromeric regions in thesepairs, while this number was reduced to two pairs in the 2n = 40F karyomorph.Homeologies among other chromosome pairs seem less reliable and must awaitfurther data. These FISH results and the presence of a unique sex chromosomesystem clearly indicate that the 2n = 40F karyomorph is a taxon with the largest suitof derived karyotypic characters in this species complex.M O L E C U L A R D ATA Molecular data are congruent with the cytogenetic results and suggest a longreproductive isolation between the karyomorphs. Haplogroup I included some ofthe 2n = 42A karyomorph haplotypes from the Tombadouro Creek and haplotypesderived from three coastal basins located south of the Para´ba do Sul River. This ıhaplogroup showed low levels of within variation, suggesting that these haplotypeswere closely related. Some haplogroup II haplotypes were sympatric with haplogroupI haplotypes at the Tombadouro Creek. Within haplogroup IIB, all haplotypes fromthe Grande River were closely related to haplotypes from four coastal basins. Thesehaplotypes showed a sister group relationship with haplotypes from the S˜ o Mateus aand Doce coastal basins that were not represented in samples from the S˜ o Francisco a© 2009 The AuthorsJournal compilation © 2009 The Fisheries Society of the British Isles, Journal of Fish Biology 2009, 75, 2326–2343
  13. 13. 2338 U. SANTOS ET AL.and Grande Rivers. Some 2n = 42 trahiras from the Tombadouro Creek and Jacar´ eRiver shared identical haplotypes, suggesting a recent range expansion from theGrande to the S˜ o Francisco basins. a These results provide insights into the evolution of H. malabaricus karyomorphsand the history of contact between coastal and continental hydrological basins.The most inclusive study of ND2 and ATPase 6 divergence between freshwaterfishes is Bermingham et al. (1997), who estimated a rate of divergence for thisgene segment of 1·3% per million years. Overall sequence (p-distance) divergenceof the divergent haplogroup I haplotypes versus haplogroup II haplotypes wouldsuggest a lineage split at 4·2 million years ago, an age that is younger than theinferred Amazon–Paranean vicariance, which has been dated at 11·8–10 million bp(reviewed in Lundberg et al., 1998). This isolation left 2n = 42 and 2n = 40 kary-omorphs in both basins (Bertollo et al., 2000), indicating that karyotypic evolutionwas well underway during the Late Miocene Epoch. Miocene fossils are also consid-ered to be similar to present-day taxa (Lundberg, 1998). The information content ofmolecular variation of this species complex may be more elusive: a preliminaryanalysis of ATPase 6 sequences in H. malabaricus populations from the Ama-zon–Paranean boundary basins reveals two clades that show large differences inp-distance divergence within each clade (0·10 and 0·34% per million years, respec-tively) (J. A. Dergam, unpubl. data), suggesting that molecular evolution within theH. malabaricus complex may be highly variable and influenced by other causes. Fac-tors such as efficiency in DNA repair mechanism (Britten, 1986), generation time(Wu & Li, 1985), metabolic rate (Martin & Palumbi, 1993) and demographic pro-cesses (Ohta, 2002) may affect substitution rates among taxa. Hence, it is plausibleto hypothesize a strong effect of demographic processes on the rates of molecularsubstitution in these populations, considering the close phylogenetic relationshipsamong H. malabaricus karyomorphs and the adaptation of H. malabaricus to life insmall populations, which is also reflected in their high levels of karyotypical vari-ation. Slightly disadvantageous substitutions can drift to high frequencies in smallpopulations (Ohta, 2002), thereby altering substitution rates in the DNA. Therefore,H. malabaricus and other species with similar ecology may be particularly poor can-didates for using the molecular clock hypothesis to estimate divergence times. In thehighly migratory species, Prochilodus spp., substitution rates of ATPase 6,8 rangefrom 0·8 to 2·5% (Sivasundar et al., 2001), suggesting haplotypes in H. malabaricusare among the most divergent in the Neotropical region. The present study is the firstto indicate such a deep molecular divergence within the same karyomorph, althoughdeep cytochrome b divergences were also observed among populations of the catfishPimelodus albicans (Valenciennes) in the River Plate basin (Vergara et al., 2008). The location where lineage splitting occurred is limited to the distribution ofcurrent karyomorphs. The Paran´ River basin harbours at least three karyomorphs, awhereas the 2n = 40F karyomorph appears to be widespread in the S˜ o Francisco aRiver basin and the 2n = 42A karyomorph is apparently restricted to the Par´ River aheadwaters. North-eastern coastal populations harbour the 2n = 40F karyomorphas far as the Buranh´ m River to the south (J. A. Dergam, unpubl. data) (Fig. 1), eand these haplotypes had the lowest degree of divergence. This macro-geographicalpattern suggests that lineage splitting occurred elsewhere and not in the relativelyisolated S˜ o Francisco River basin. In haplogroup IIB, the close relationship between athe 2n = 42A and 2n = 42B karyomorphs indicated that karyotypical differentiation © 2009 The Authors Journal compilation © 2009 The Fisheries Society of the British Isles, Journal of Fish Biology 2009, 75, 2326–2343
  14. 14. PHYLOGEOGRAPHY OF HOPLIAS MALABARICUS 2339involved few alterations that resulted in a differentiated XX/XY chromosome system(Bertollo et al., 2000).S T R E A M P I R AC Y On a local scale, the apparently restricted range of the Tombadouro Creek 2n =42A karyomorph may have resulted from stream piracy from a former Jacar´ River etributary. The Tombadouro Creek is 12 km from the Jacar´ across the Galga water- eshed. Three hypotheses might explain this dispersal. First, boundary crossing mayhave resulted from either human mediated activities. Second, dispersal may haveoccurred through a high-altitude wetland that used to drain into both basins. Forexample, the connection may have been destroyed by the construction of BR 494Highway, which is 8 km away at its closest point from the Tombadouro Creek. Third,headwater capture in the region may have occurred by relatively recent reactivationof ancient faults (Saadi et al., 2002). Altitudinal characteristics are also consistentwith the direction of stream capture. Tombadouro is at 860 m altitude, whereas theJacar´ River is at 1038 m. This area is within the range of a continental fault known eas the Upper Grande River Crustal Discontinuity (Saadi et al., 2002). This faultrepresents the southern limit of the S˜ o Francisco craton and the movable belts of asouthern Minas Gerais State (Fig. 1) and has been active in the last 15 000 years. The phylogenetic origins of two highly divergent haplotypes in the TombadouroCreek became clear only after haplotypes from coastal basins were included in theanalysis. Most critically, haplotypes found in sympatry in the S˜ o Francisco River aare related to haplotypes in different coastal basins. Under the allopatric model ofspeciation, this asymmetry suggests that ancestral fish dispersed from the coastal tothe continental drainages, and the present study is the first to indicate unambiguouslythe direction of this dispersal. The timing of the dispersal between coastal and inlanddrainages, however, could not be estimated, because molecular clock expectations arenot suitable for H. malabaricus. Therefore, the phylogeographic history of this groupdoes not fit any of the three phylogenetic patterns proposed by Ribeiro (2006). On theother hand, low genetic divergence between the S˜ o Francisco and Itapicuru River adrainages may be part of a recent faunal exchange in the arid region of north-easternBrazil that is apparent in many other species of fish (Rosa et al., 2004). Ribeiro (2006) indicated that the Atlantic drainages (Doce, Para´ba do Sul and ıRibeira) have existed since the break-up of Gondwana. The results of the presentstudy suggested that populations in these basins represent three biogeographic units,for which only two have close representatives in the boundaries of the Jacar´ and Par´ e adrainages. Dergam et al. (2002) previously reported a close phylogenetic relatednessbetween two haplotypes in the Doce and lower Grande River basins. Fish dispersalfrom coastal to continental drainages may have been restricted to only a few taxaadapted to headwater conditions, because coastal drainages are characterized by highlevels of endemism (Ribeiro, 2006) and the Proterozoic Mantiqueira Range is con-sidered to be an efficient barrier between coastal and continental drainages (Ingenito& Buckup, 2007). To the south, stream piracy has been particularly intensive in thePara´ba do Sul, Ribeira and Upper Paran´ basins (Ribeiro, 2006). ı a The present study provided important information on the population distributionof H. malabaricus in S˜ o Francisco River basin and on historical relationships with apopulations in the Grande River and coastal basins. The 2n = 40F karyomorph and© 2009 The AuthorsJournal compilation © 2009 The Fisheries Society of the British Isles, Journal of Fish Biology 2009, 75, 2326–2343
  15. 15. 2340 U. SANTOS ET AL.its associated haplotypes were widely distributed throughout the S˜ o Francisco basin, amaintaining an apparently stable karyotypic structure of the chromosomal markersanalysed here. This karyomorph was characterized by a XX/XY sex chromosomesystem. In contrast, the 2n = 42A karyomorph showed a more restricted distributionin S˜ o Francisco basin and was closely associated with populations in a Grande aRiver basin tributary, where the 2n = 42A karyomorph predominated. The 2n = 42Akaryomorph populations isolated in drainages, however, showed striking differencesfor some markers, notably the Ag-NORs and 5S rDNA sites. The patterns of variationin repetitive DNA sequences as revealed by 5S rDNA probes indicated that thesemarkers appeared to be useful population markers, showing significant differencesamong localities and karyomorphs. Patterns of 5SHind III-DNA variation yielded thelargest phylogenetic signal and placed the 2n = 40F karyomorph in a high positionin the phylogeny of karyomorphs that occur in eastern Brazil. Given the current status of phylogenetic data, an understanding of distributionpatterns for H. malabaricus will be particularly challenging. Studies of the distri-bution patterns of freshwater fishes have considered endemism (Vari, 1988), faunalcomposition including endemic and widespread species (Hubert & Renno, 2006),molecular data (Sivasundar et al., 2001; Montoya-Burgos, 2003) or a combinationof morphological and molecular data (Menezes et al., 2008). When integrated into awider geographic context, molecular and cytogenetic patterns can be informative offaunal dispersals between drainages. The results of the present study suggested thatrecent fish dispersal may occur even within apparently stable geological regions andthat headwater stream piracy between coastal and continental (upper Paran´ tribu- ataries) basins may have played a role in producing the high levels of freshwater fishdiversity in the Neotropics. Finally, the topology of mtDNA-based tree suggests that the proposal of Dergam &Bertollo (1990) to separate species by diploid numbers would result in a paraphyleticarrangement, in which some Hoplias spp. haplotypes associated with 2n = 42 kary-omorphs would be more closely related to trahiras with 2n = 40 karyomorphs thanto other fish with 2n = 42 karyomorphs. Although haplogroup I appeared as the sis-ter group of the remaining trahiras, haplogroup II is less supported and the possibleancestry of karyomorphs must wait for more data. A time frame for dispersal betweencoastal and continental drainages is elusive for the H. malabaricus species complex.Nevertheless, geographical isolation during coastal diversification has enhanced spe-ciation processes and genetic isolation between karyomorphs, a process that hasresulted in molecular and cytogenetic differentiation between populations. Theseresults are consistent with the existence of multiple independent lineages that can beeasily detected with standard cytogenetic techniques. Although Hoplias spp. are common members of lowland freshwater communities(Bonetto et al., 1969; Winemiller, 1991), at least part of their widespread distributionin the Neotropics is due to their ability to colonise high-altitude headwaters, suchas in the Jacar´ and Tombadouro localities. The question of how this non-migratory esedentary species has adapted to sluggish waters and to waters at high altitudes is notexplained by current geological data. Nevertheless, this broad ecological diversityindicates a long history of Hoplias spp. in south-eastern Brazil. J. T. da Costa, J. B. Santos, R. de F. Guimar˜ es and A. M. R. de Souza provided support aduring fieldwork. W. Evangelista and U. Jacobina collected samples from the Ibimirim lakes © 2009 The Authors Journal compilation © 2009 The Fisheries Society of the British Isles, Journal of Fish Biology 2009, 75, 2326–2343
  16. 16. PHYLOGEOGRAPHY OF HOPLIAS MALABARICUS 2341and Itapicuru River, and T. L. Pereira assisted with some laboratory work. M. Wachlevskisupported the initial stage of this study. The authors wish to thank two anonymous review-ers for their suggestions. Financial support for this study was provided by the Program forthe Advancement of Research of the Universidade do Estado de Minas Gerais, the NationalCouncil for Technological and Scientific Development of the Brazilian Government, the Uni-versidade Federal de Vicosa and the Fundacao Educacional de Divin´ polis-FUNEDI/UEMG. ¸ ¸˜ o ReferencesBermingham, E., McCafferty, S. & Martin, A. P. (1997). Fish biogeography and molecular clocks: perspectives from the Panamanian Isthmus. In Molecular Systematics of Fishes (Kocher, T. D. & Stepien, C. A., eds.), pp. 113–128. San Diego: Academic Press.Berra, T. M. (2007). Freshwater Fish Distribution. Chicago, IL: University of Chicago Press.Bertollo, L. A. C. (1996). The nucleolar organizer regions of Erythrinidae fish. An uncommon situation in the genus Hoplias. Cytologia 61, 75–81.Bertollo, L. A. C., Born, G. G., Dergam, J. A., Fenocchio, A. S. & Moreira-Filho, O. (2000). A biodiversity approach in the neotropical Erythrinidae fish, Hoplias malabaricus. Karyotypic survey, geographic distribution of cytotypes and cytotaxonomic considerations. Chromosome Research 8, 603–613.Bertollo, L. A. C., Moreira-Filho, O. & Fontes, M. S. (1997). Karyotypic diversity and distribution in Hoplias malabaricus (Pisces, Erythrinidae): cytotypes with 2n = 40 chromosomes. Genetics and Molecular Biology 20, 237–242.Bertollo, L. A. C., Moreira-Filho, O. & Galetti, P. M. Jr. (1986). Cytogenetics and taxonomy considerations based on chromosome studies of freshwater fish. Journal of Fish Biology 28, 153–159.Bertollo, L. A. C., Takahashi, C. & Moreira-Filho, O. (1978). Cytotaxonomic considerations on Hoplias lacerdae (Pisces, Erythrinidae). Brazilian Journal of Genetics 1, 103–120.Britten, R. J. (1986). Rates of DNA sequence evolution differ between taxonomic groups. Science 231, 1393–1398.Bonetto, A., Dioni, W. & Pignalberi, C. (1969). Limnological investigations on biotic com- munities in the Middle Paran´ River Valley. Verhandlungen Internationalen Vereinigung a f¨ r Theoretische und Angewandte Limnologie 17, 1035–1050. uBorn, G. & Bertollo, L. A. C. (2000). An XX/XY sex chromosome system in a fish species, Hoplias malabaricus, with a polymorphic NOR-bearing X chromosome. Chromosome Research 8, 111–118.Born, G. & Bertollo, L. A. C. (2001). Comparative cytogenetics among allopatric populations of the fish, Hoplias malabaricus. Cytotypes with 2n = 42 chromosomes. Genetics and Molecular Biology 110, 1–9.Boyce, T. M., Zwick, M. E. & Aquadro, C. F. (1989). Mitochondrial DNA in the bark weevils: size, structure and heteroplasmy. Genetics 123, 825–836.Cioffi, M. B., Martins, C. & Bertollo, L. A. C. (2009). Comparative chromosome mapping of repetitive sequences. Implications for genomic evolution in the fish, Hoplias malabaricus. BMC Genetics 10, 34. doi: 10.1186/1471-2156-10-34Dergam, J. A. & Bertollo, L. A. C. (1990). Karyotypic diversification in Hoplias malabaricus (Osteichthyes, Erythrinidae) of the S˜ o Francisco and Alto Paran´ basins, a a Brazil. Genetics and Molecular Biology 13, 755–756.Dergam, J. A., Suzuki, H. I., Shibatta, O. A., Duboc, L. F., J´ lio, H. F., Caetano-Giuliano, L. u & Black, W. C. I. V. (1998). Molecular biogeography of the Neotropical fish Hoplias malabaricus (Erythrinidae: Characiformes) in the Iguacu, Tibagi, and Paran´ rivers. ¸ a Genetics and Molecular Biology 21, 493–496.Dergam, J. A., Paiva, S. R., Schaeffer, C. E., Godinho, A. L. & Vieira, F. (2002). Phylo- geography and RAPD-PCR variation in Hoplias malabaricus (Bloch,1794) (Pisces, Teleostei) in southeastern Brazil. Genetics and Molecular Biology 25, 379–387.Felsenstein, J. (1981). Evolutionary trees from DNA sequences: a maximum likelihood approach. Journal of Molecular Evolution 17, 368–376.Ferreira, I. A., Bertollo, L. A. C. & Martins, C. (2007). Comparative chromosome mapping of 5SrDNA and 5S Hin dIII repetitive sequences in Erythrinidae fishes (Characiformes)© 2009 The AuthorsJournal compilation © 2009 The Fisheries Society of the British Isles, Journal of Fish Biology 2009, 75, 2326–2343
  17. 17. 2342 U. SANTOS ET AL. with emphasis on the Hoplias malabaricus ‘species complex’. Cytogenetic and Genome Research 118, 78–83.Haaf, T., Schmid, M., Steinlein, C., Galetti, P. M. Jr & Willard, H. F. (1993). Organization and molecular cytogenetics of satellite DNA family from Hoplias malabaricus (Pisces, Erythrinidae). Chromosome Research 1, 77–86.Harvey, P. H. & Pagel, M. D. (1991). The comparative method in evolutionary biology. Oxford Series in Ecology and Evolution. p. 263. New York, NY: Oxford University Press.Henyey, E., Kynard, B. & Zhuang, P. (2002). Use of eletronarcosis to immobilize juvenile lake and shortnose sturgeons for handling and the effects on their behavior. Journal of Applied Ichthyology 18, 502–504.Howell, W. M. & Black, D. A. (1980). Controlled silver staining of nucleolus organizer regions with a protective colloidal developer: a 1-step method. Experientia 36, 1014–1015.Hubert, N. & Renno, J. F. (2006). Historical biogeography of South American freshwater fishes. Journal of Biogeography 33, 1414–1436.Huelsenbeck, J. P. & Ronquist, F. (2001). MrBayes: Bayesian inference of phylogeny. Bioin- formatics 17, 754–755.Ingenito, L. F. S. & Buckup, P. A. (2007). The Serra da Mantiqueira, south-eastern Brazil, as a biogeographical barrier for fishes. Journal of Biogeography 34, 1173–1182.Levan, A., Fredga, K. & Sandberg, A. A. (1964). Nomenclature for centromeric position on chromosome. Hereditas 1, 201–220.Lundberg, J. G., Marshall, L. G., Guerrero, J., Horton, B., Malabarba, M. C. S. L. & Wesselingh, F. (1998). The stage for Neotropical fish diversification: a History of tropical South American rivers. In Phylogeny and Classification of Neotropical Fishes (Malabarba, L. R., Reis, R. E., Vari, R. P., Lucena, C. A. S. & Lucena, Z. M. S., eds), pp. 13–48. Porto Alegre, Brazil: Edipucrs.Lundberg, J. G. (1998). The temporal context for the diversification of Neotropical fishes. In Phylogeny and Classification of Neotropical Fishes (Malabarba, L. R., Reis, R. E., Vari, R. P., Lucena, C. A. S. & Lucena, Z. M. S., eds), pp. 49–68. Porto Alegre, Brazil: Edipucrs.Martin, A. P. & Palumbi, S. R. (1993). Body size, metabolic rate, generation time, and the molecular clock. Proceedings of the National Academy of Sciences of the United States of America 90, 4087–4091.Martins, C., Ferreira, I. A., Oliveira, C., Foresti, F. & Galetti, P. M. Jr. (2006). A tandemly repetitive centromeric DNA sequence of the fish Hoplias malabaricus (Characiformes: Erythrinidae) is derived from 5S rDNA. Genetica 127, 133–141.Menezes, N. A., Ribeiro, A. C., Weitzman, S. & Torres, R. A. (2008). Biogeography of Glan- dulocaudinae (Teleostei: Characiformes: Characidae) revisited: phylogenetic patterns, historical geology and genetic connectivity. Zootaxa 1726, 33–48.Molina, W. F. (2001). An alternative method of mitotic stimulation in fish cytogenetics. Chromosome Science 5, 149–152.Montoya-Burgos, J. I. (2003). Historical biogeography of the catfish genus Hypostomus (Siluriformes: Loricariidae), with implications on the diversification of Neotropical ichthyofauna. Molecular Ecology 12, 1855–1867.Myers, G. S. (1938). Fresh-water fishes and the West Indian zoogeography. Annual Report of the Smithsonian Institution 1937, 339–364.Nylander, J. A. A. (2004). MrModeltest v2. Program distributed by the author. Uppsala, Sweden: Evolutionary Biology Centre, Uppsala University.Ohta, T. (2002). Near neutrality in evolution of genes and gene regulation. Proceedings of the National Academy of Sciences of the United States of America 99, 16134–16137.Pinkel, D., Straume, T. & Gray, J. (1986). Cytogenetic analysis using quantitative, high sen- sitivity, fluorescence hybridization. Proceedings of the National Academy of Sciences of the United States of America 83, 2934–2938.Posada, D. & Crandall, K. A. (1998). Modeltest: testing the model of DNA substitution. Bioinformatics 14, 817–818. © 2009 The Authors Journal compilation © 2009 The Fisheries Society of the British Isles, Journal of Fish Biology 2009, 75, 2326–2343
  18. 18. PHYLOGEOGRAPHY OF HOPLIAS MALABARICUS 2343Quenouille, B., Bermingham, E. & Planes, S. (2004). Molecular systematics of the dam- selfishes (Teleostei: Pomacentridae): Bayesian phylogenetic analyses of mitochondrial and nuclear DNA sequences. Molecular Phylogenetics and Evolution 31, 66–88.Reis, R. E., Kullander, S. O., & Ferraris, C. J. Jr. (2003). Check List of the Freshwater Fishes of South and Central America. Porto Alegre, Brazil: Edipucrs.Ribeiro, A. C. (2006). Tectonic history and the biogeography of the freshwater fishes from the coastal drainages of eastern Brazil: an example of faunal evolution associated with a divergent continental margin. Neotropical Ichthyology 4, 225–246.Rosa, R. S., Menezes, N. A., Britski, H. A., Costa, W. J. E. M. & Groth, F. (2004). Diver- sidade, padr˜ es de distribuicao e conservacao dos peixes da caatinga. In Ecologia e o ¸˜ ¸˜ conserva¸ ao da Caatinga (Tabarelli, I. R. M. & da Silva, J. M. C., eds), pp. 135–180. c˜ Recife: Edufpe.Saadi, A., Dart, R. L. & Machette, M. N. (2002). Map of quaternary faults and lineaments of Brazil. Project of international lithosphere program task group II-2, major active faults of the world, University of Minas Gerais and U.S. Geological Survey.Saitou, N. & Nei, M. (1987). The neighbor-joining method: a new method for reconstructing phylogenetic trees. Molecular Biology and Evolution 4, 406–425.Schaefer, S. A. (1998). Conflict and resolution: Impact of new taxa on phylogenetic studies of the neotropical cascudinhos (Siluriformes: Loricariidae). In Phylogeny and Classifica- tion of Neotropical Fishes (Malabarba, L. R., Reis, R. E., Vari, R. P., Lucena, C. A. S. & Lucena, Z. M. S., eds), pp. 375–400. Porto Alegre, Brazil: Edipucrs.Sites, J. W. Jr & Moritz, C. (1987). Chromosomal evolution and speciation revisited. System- atic Zoology 36, 153–174.Sivasundar, A., Bermingham, E. & Ort´, G. (2001). Population structure and biogeography ı of migratory freshwater fishes (Prochilodus:Characiformes) in major South American rivers. Molecular Ecology 10, 407–417.Sola, L., Rossi, A. R., Laselli, V., Rasch, E. M. & Monaco, P. J. (1992). Cytogenetics of bisexual/unisexual species of Poecilia II. Analysis of heterochromatin and nucleolar organizer regions in Poecilia mexicana mexicana by C-banding and DAPI, quinacrine, chromomycin A3, and silver staining. Cytogenetics and Cell Genetics 60, 229–235.Sumner, A. T. (1972). A simple technique for demonstrating centromeric heterocromatin. Experimental Cell Research 75, 304–306.Tamura, K., Dudley, J., Nei, M. & Kumar, S. (2007). MEGA4: molecular evolutionary genetics analysis (MEGA) software version 4.0. Molecular Biology and Evolution 24, 1596–1599.Vari, R. P. (1988). The Curimatidae, a lowland Neotropical fish family (Pisces: Characi- formes); distribution, endemism, and phylogenetic history. In Proceedings of a Workshop on Neotropical Distribution Patterns. (Heyer, W. R. & Vanzolini, P. E., eds), pp. 343–377. Rio de Janeiro: Academia Brasileira de Ciˆ ncias. eVergara, J., Azpelicueta, M. M. & Garcia, G. (2008). Phylogeography of the Neotropical cat- fish Pimelodus albicans (Siluriformes: Pimelodidae) from R´o de la Plata Basin, South ı America, and conservation remarks. Neotropical Ichthyology 6, 75–85.Vicari, M. R., Artoni, R. F. & Bertollo, L. A. C. (2003). Heterochromatin polymorphism associated with 18S rDNA. A differential pathway among the fish Hoplias malabaricus from Southern Brazil. Cytogenetic and Genome Research 101, 24–28.Vicari, M. R., Artoni, R. F. & Bertollo, L. A. C. (2005). Comparative cytogenetics of Hoplias malabaricus (Pisces, Erythrinidae): a population analysis in adjacent hydrographic basins. Genetics and Molecular Biology 28, 103–110.Winemiller, K. O. (1991). Ecomorphological diversification in lowland freshwater fish assemblages from five biotic regions. Ecological Monographs 61, 343–365.Wu, C. I. & Li, W. H. (1985). Evidence for higher rates of nucleotide substitution in rodents than in man. Proceedings of the National Academy of Sciences of the United States of America 82, 1741–1745.© 2009 The AuthorsJournal compilation © 2009 The Fisheries Society of the British Isles, Journal of Fish Biology 2009, 75, 2326–2343